from Follow the Money

You can not put dark matter in a container …

April 19, 2006

Blog Post

More on:

United States

Budget, Debt, and Deficits


So says Michael Mandel.  

Some of us might say you can't put dark matter in a container because it doesn't really exist.

That is not Mandel's argument.  He believes in dark matter, though apparently not in the container revolution.

Call me old economy.   Like William Polley, I suspect that containerization was a big deal.   Without the efficiencies created from factories in China to container ships to Long Beach to distribution centers to trucks to big boxes that Americans willingly drive miles to get too, lots of goods would cost more.  Walmart - one of Mandel's favorite companies - wouldn't be Walmart without the container.

Actually, call me old economy and old accounting.   I don't buy the case for dark matter.  

Not after reading Daniel Gros.

Daniel Gros notes that if you believe in dark matter, you have to believe that the US has systematically taken foreign investors to the cleaners.

Remember, dark matter arises not because the return on US foreign direct investment is particularly good, but because the return on foreign direct investment in the US is terrible.   And since most FDI is the US comes from Europe and Canada, the US data suggests that Europeans - including our Thatcherite friends in Britain -- kan't manage.

If you believe in dark matter, you believe:

  • That the return on European investment in the US dramatically falls the moment European equity investors take a stake bigger than 10% in a US company.   Gros: "The official data imply that the foreigners instantly start losing when they invest more than 10% into a US company ... it seems that foreign investors in the US are able to obtain a market rate of return if they own less than 10% of a US company.   But the official statistics imply that their returns are much lower once the investment qualifies as direct."
  • European firms invest in US factories when they would be better off investing in US treasuries.  Not on a risk-adjusted basis.  Absolutely.  They go to all the trouble of managing US firms to earn less than they get on Treasuries.  Gross: "if the official statistics were correct, foreigners would accept a rate of return on their direct investment in the US of only around 2.5%, which would be much lower not only of that on their US portfolio equity investment, but also lower than the 5.5% return earned on debt instruments."
  • The equity risk premium for foreign FDI in the US is negative. European foreign investors happily accept lower returns on their FDI to avoid the risks associated with US bonds:  "The official data based on accounting returns would thus imply a strongly negative equity risk premium (for FDI alone)."

If you don't think any of this makes sense, then you have to believe that US data mismeasures either US returns on its foreign direct investment abroad or returns on foreign direct investment in the US.

In the past, I - and PGL over at Angry Bear - have suggested that US and European firms manipulate transfer pricing to shift profits to low tax European jurisdictions.   There is some evidence that is true.   Philip Lane and Frances Ruane should know a thing or two about this kind of Green matter.  Philip Lane is from Ireland, the preferred tax haven of American software firms and big Pharma. And he and Ruane found that the majority of FDI in Ireland comes from the US, and that it earns a return of about 17.5% ...  

"The high yields earned on FDI in Ireland may in part be attributable to an under-estimation of the value of foreign-owned capital in Ireland or to an over-estimation of the profitability of their operations.  The latter may be the result of the tax planning activities of multinational corporations."

(quote is on p.20 of the Lane/ Ruane paper)

It sure seems that US firms operating in the intangible economy like to be taxed at the low Irish rate than the high US rate ...

But Daniel Gros makes a different argument.   He argues that a normal return on FDI implies that foreigners should be earning around $220b in the US, not the $100-120b reported in the US data as recently as 2004.  And he doesn't think there is any evidence transfer pricing has reduced the (reported) return on foreign direct investment in the US by $100 billion. 

Rather, he thinks the low apparent return on FDI in the US stems from differences in the way reinvested earnings are reported.

A fair amount of the "profits" that show up in the US balance of payments data come from reinvested earnings.   Technically, reinvested earnings should enter as a credit in the current account, and as an offsetting outflow in the capital account.  US firms effectively finance an increase in their investment abroad out of some of their earnings abroad.

Fair enough.   It is a consistent way of treating reinvested earnings in the international statistics.

But it turns out that US firms report their reinvested earnings to the US data collectors (no need to worry about a tax bill, the US doesn't tax earnings held abroad), but foreign investors in the US don't report any reinvested earnings.   Hmmm.

A couple of data points Gros highlights:

  • Between 1982 and 2004, reinvested earnings on US FDI abroad amounted to $1,100b, while foreign investors in the US have only reinvested $20 billion over the same period.
  • From 1999-2004, US firms reported an average of $103-104b in reinvested earnings, while foreigners reported only $7b a year.

Consequently, the average difference between the "distributed" earnings on US FDI abroad and foreign investment in the US was only $23b, while the average difference between reinvested earnings was $95b.

I'll translate that into dark matter.   The difference in cash earnings is the source of $460b in dark matter - using the Hausmann and Sturzenegger assumption that cash flows should be discounted at 5%.  The difference in retained earnings, by contrast, accounted for $1920b in "dark matter."

That is real (dark) money.

If you believe that European firms in the US don't reinvest any of their earnings, the US BoP data is accurate, and the US income balance is more or less in surplus.   But if you believe that European firms do have some retained earnings, but they don't appear in the US data, then the income data - the key to the discovery of dark matter -- has some problems.

Gros thinks the US should either add $100b in estimated retained earnings on FDI in the US, or subtract retained earnings on US FDI from the income balance (Gros' preferred solution).

That would increase the 2004 US current account deficit from $668 b to $788b - and, one assumes, push the 2005 current account deficit above $900b.  

The fact that the US data is not capturing retained earnings also implies that the stock of FDI in the US is underestimated in the official data, and the United States net international investment position is worse than it seems in the official data.  Gros:

"A more accurate estimate of the value of US direct investment abroad also implies that the widely shared view that the US operates like a hedge fund, i.e. short on debt and long on equity, especially FDI, is simply wrong.   The US has already reached a large debtor position in terms of FDI.   ... Another implication of the correction is that that US GDP is being overstated (relative to GNP) by about 1%, since the returns on US FDI abroad are overestimated relative to the returns foreign firms earn on their direct investment in the US. "

Remember, that in countries with lots of external debt and foreign direct investment, the country produces more than it gets, since foreigners get the profits on all the investment that they have financed.

Gros also thinks the US data undercounts foreign holdings of US debt.   He notes that valuation gains on US investment abroad cannot explain the difference between the sum of past US current account deficits and the stock of claims on the US.   The big gains the US enjoyed in recent years as a falling dollar (rising euro) increased the dollar value of US investment abroad largely offset large valuation losses in the late 1990s, when the rising dollar (falling euro) reduced the dollar value of US investment in Europe.  

So why hasn't the United States net indebtedness increased by more?   In his view, foreigners systematically underreport their holdings of US debt in the US Treasury surveys.

"The US has acted like a black hole for capital from the rest of the world: one can observe a large amount of investment flowing into the US, but after some time it disappears from the statistics (and foreign investment in the US that takes the form of FDI earns no returns).  The discrepancy arises for a simple reason: the flow data are based on actual flows of payments recorded in the balance of payments.  By contrast, the stock data are based on US surveys, which tend to miss out on US assets held by foreigners."

That would explain something I have noticed as well.  When the NIIP is revised, it always seems to be revised downward.   

Gross notes that the estimated net debt position of the US at the end of 2000 was estimated at negative $2.2 trillion in 2001, and only negative 1.6 trillion at the end of 2005.  Why?  The estimated value of US FDI abroad grew, and, more puzzling, the estimated stock of foreign holdings of US debt FELL.   Why?  Gros: "The most likely reason for the fall in reported holdings of corporate bonds is probably simply that these bonds were not held by US custodians (to avoid tax on the interest income?)"

Bottom lines:

  • The high returns on US investment in Ireland sure look like tax arbitrage.
  • The low reported return on foreign direct investment in the US makes no economic sense. 
  • The fact that the US data shows that US firms investing abroad have large reinvested earnings while foreign firms investing in the US have no reinvested earnings doesn't make much sense either.    After you adjust for this fact, a lot of the  "dark matter" that Hausmann and Sturzenegger claim to have discovered suddenly disappears.  
  • Tax arbitrage may reduce the reported stock of debt claims on the US, not just the reported income on foreign assets on foreign direct investment in the US.
  • According to Gros, the United States' real external position is far worse than the official data reports.  The US has more net external debt.   The US income balance is already in the red.  And the US current account deficit is BIGGER than officially reported.

The US had better hope that US creditors in the Gulf and Asia believe our Latin friends (Hausmann and Sturzenegger), not the grunts cranking through the data over in Europe ...

More on:

United States

Budget, Debt, and Deficits